This study estimates the effects of government spending using local projection methods, with a focus on the crowding out effect of different types of government expenditures. I find that government spending on consumption non-durables crowds out private consumption and government capital expenditures crowd out both private consumption and investment. Current transfers crowd out private consumption and investment, and transfers that subsidize capital use crowd out private investment with a lag. Differences in the intertemporal elasticity of substitution, marginal cost of production, markups,and wealth effects contribute to the different effects of government spending by expenditure characteristics.